Universities and the Development of Industry ClustersExecutive Summary Regional efforts to develop industry clusters increasingly include universities as central assets. Unfortunately,

Universities and the Development of Industry Clusters

Center for Economic Development

Carnegie Mellon UniversityCenter for Economic DevelopmentUTDC Suite 2084516 Henry StreetPittsburgh, PA 15213

Phone: 412.268.9880Fax: 412.268.9828www.smartpolicy.org

Smart Policy for Innovative Regions

2 0 0 4

Carnegie MellonHeinz SchoolPolicy • Management •Information Technology

ECONOMIC DEVELOPMENT ADMINISTRATIONU.S. Department of Commerce

ECONOMIC DEVELOPMENT ADMINISTRATION

Universities and the Development of Industry Clusters

Jerry Paytas, Ph.D., Robert Gradeck and Lena Andrews.

Carnegie Mellon University, Center for Economic Development

With the assistance of Alexis Haakensen, Anjani Datla, Timothy Collins, and David Anderson.

2004

Prepared for

Economic Development Administration

U.S. Department of Commerce

This report was prepared by the Carnegie Mellon Center for Economic Development under award number 99-07-13822 from the Economic Development Administration, U.S. Department of Commerce, and grant agreement number B0125 from the Vira I. Heinz Endowment. The findings and conclusions are those of the authors alone and do not reflect the views or policies of the Economic Development Administration, the U.S. Department of Commerce, Carnegie Mellon University, or any officers thereof.

jp87Cover design by Gary Franko

A Message from David A. Sampson, Assistant Secretary for Economic Development, U.S. Department of Commerce

The bottom-line of economic development today is about building prosperity – a high and rising standard of living. Productivity and productivity growth are the fundamental drivers of prosperity and innovation is the key driver of productivity. The focus of economic development should be on supporting innovation, increasing prosperity for American businesses and ensuring American workers have the skills to remain the most productive workforce in the world. Innovation will drive the growth of American industry by fostering new ideas, technologies and processes that lead to better jobs and higher wages – and as a result, a higher standard of living. America’s capacity to innovate will serve as its most critical element in sustaining economic growth.

The dominant reality of economic development today is that we live and operate in a worldwide economy. Worldwide commerce means that American businesses must operate and cooperate with countries around the world. Consequently, we must think regionally, avoid isolationist practices and build a strong economic platform for growth.

Over the course of its history, EDA has encouraged communities to harness the wealth of intellectual and technical resources of institutions of higher education to promote regional economic development. Universities can be a powerful force when they effectively serve the development needs of local communities. This research focuses on the key role universities can play in cluster-based economic development, establishes a foundation for a critical review of that role, and assesses the factors that are vital to successful university-industry cluster development.

David A. Sampson

Table of Contents Executive Summary ............................................................................................................. i Introduction......................................................................................................................... 1 Findings............................................................................................................................... 2

University Factors........................................................................................................... 4 Breadth of Involvement .............................................................................................. 8 Strong Base of Research ........................................................................................... 11 Regional Alignment .................................................................................................. 13

Cluster Factors .............................................................................................................. 21 Networked Cluster (Industrial Districts)................................................................... 24 Hub and Spoke.......................................................................................................... 26 Satellite (Branch Plant) ............................................................................................. 27 Institutional Clusters (State-Anchored) .................................................................... 31

Conclusion .................................................................................................................... 34 Case Profiles ..................................................................................................................... 35

University of Michigan / Ann Arbor ............................................................................ 36 Wright State University / Dayton ................................................................................. 44 New Mexico State University / Las Cruces.................................................................. 51 Lehigh University / Lehigh Valley (Allentown)........................................................... 57 West Virginia University / Morgantown ...................................................................... 66 Virginia Polytechnic University (Virginia Tech)/ New River Valley .......................... 73 University of Northern Iowa / Waterloo-Cedar Falls ................................................... 82 Florida State University /Tallahassee ........................................................................... 88

Appendix 1 – Case Selection Methodology...................................................................... 95 Initial Case Selection ................................................................................................ 95 Assessing University Involvement ........................................................................... 98

Appendix 2 – Incentive Programs in the Case States ..................................................... 100 Appendix 3 – Data Sources for Statistics ....................................................................... 102

List of Tables Table 1: Case Taxonomy .................................................................................................... 1 Table 2: Tech Transfer in Case Study Universities ........................................................... 6 Table 3: University-Cluster Alignment ........................................................................... 14 Table 4: Sample FSU Cluster Assets ................................................................................ 20 Table 5: Technology Firms in Tallahassee ....................................................................... 21 Table 6: Cluster Typology and Cases .............................................................................. 24 Table 7: Employment in Transportation Equipment ....................................................... 28 Table 8: Transportation Equipment ................................................................................. 29 Table 9: Ann Arbor Economic Statistics ......................................................................... 37 Table 10: University of Michigan Significant Degree Categories................................... 37 Table 11: Dayton Economic Statistics............................................................................. 45 Table 12: Wright State Significant Degree Categories..................................................... 45 Table 13: Las Cruces Economic Statistics....................................................................... 52

Table 14: New Mexico State University Significant Degree Categories ......................... 53 Table 15: Lehigh Valley Economic Statistics.................................................................. 58 Table 16: Lehigh University Significant Degree Categories........................................... 59 Table 17: Monongalia County Economic Statistics......................................................... 67 Table 18: West Virginia University Significant Degree Categories................................. 68 Table 19: New River Valley Economic Statistics............................................................. 74 Table 20: Virginia Tech Significant Degree Categories................................................... 74 Table 21: Cedar Valley Economic Statistics .................................................................... 82 Table 22: University of Northern Iowa Significant Degree Categories .......................... 83 Table 23: Tallahassee Economic Statistics ....................................................................... 89 Table 24: Florida State University Significant Degree Categories................................... 89 Table 25: Selection Criteria ............................................................................................. 95 Table 26: Initial Case Selection ....................................................................................... 97 Table 27: Factors for Engaged Universities..................................................................... 98 Table 28: Engaged Universities ....................................................................................... 99 Table 29: Methods of Financing.................................................................................... 100 Table 30: Geographic Targeting .................................................................................... 101

List of Figures Figure 1: Case Taxonomy................................................................................................... 2 Figure 2: Cluster Policy ...................................................................................................... 3 Figure 3: Models for Organizing University Tech Transfer Activity................................. 5 Figure 4: Annual R&D Spending and Start Ups............................................................... 12 Figure 5: Start Ups and R&D Expenditures, 1998-2000 .................................................. 13 Figure 6: University of Michigan Alignment ................................................................... 15 Figure 7: Lehigh Alignment.............................................................................................. 16 Figure 8: Wright State Alignment..................................................................................... 18 Figure 9: Florida State University Alignment .................................................................. 19 Figure 10: Networked Clusters ........................................................................................ 24 Figure 11: Hub and Spoke Clusters .................................................................................. 27 Figure 12: Satellite Clusters.............................................................................................. 27 Figure 13: Key Clusters in the New River Valley ........................................................... 30 Figure 14: Institutional Clusters........................................................................................ 31 Figure 15: University of Michigan R&D by Discipline, 2000 ......................................... 39 Figure 16: Wright State R&D by Discipline, 2000........................................................... 46 Figure 17: New Mexico State R&D by Discipline, 2000 ................................................. 54 Figure 18: Lehigh University R&D by Discipline, 2000.................................................. 61 Figure 19: West Virginia University R&D by Discipline, 2000 ...................................... 69 Figure 20: Virginia Tech R&D by Discipline, 2000......................................................... 75 Figure 21: Key Clusters in the New River Valley ............................................................ 79 Figure 22: University of Northern Iowa R&D by Discipline, 2000 ................................. 84 Figure 23: Universities with Highest Involvement in Regional Economies..................... 96 Figure 24: State Incentive Programs by Finance Type................................................... 100

Executive Summary

Regional efforts to develop industry clusters increasingly include universities as central assets. Unfortunately, little is understood about how universities impact the development of regional industry clusters. Practice often precedes policy and such is the case with university- and cluster-based strategies. States and regions have jumped on the bandwagon before there is established knowledge of the costs and benefits of these approaches. This research establishes a foundation for a critical review of the role of the university in cluster development. It assesses both the factors of the university and the factors of the cluster that are vital to successful university-industry cluster development.

Universities that are highly engaged with regional industry clusters have diverse and complementary units that broadly address the needs of the cluster. Rather than a compartmentalized approach, engaged universities are sources of research and technology, but also address other aspects that affect cluster growth such as business, marketing, legal, and workforce issues. In order to have an impact on a regional industry cluster, the university must have a significant base of research aligned with the needs of that cluster. In the case of research and technology assets, size does matter. The university must have a large base of research and development in order to significantly impact a cluster, rather narrowly benefiting only a few firms. The university must also have expertise and resources in appropriate areas that align with the needs of the clusters in the region. Less important is the structure or processes of the technology transfer function. The key factors for universities, discussed in the findings are:

• Breadth of involvement

• Strong base of R&D

• Regional alignment

The characteristics of the cluster are as important as the characteristics of the university if there is to be any regional impact. Universities cannot defy the forces of the market. Established clusters with mature products and processes are less receptive to innovation, especially from universities and other external sources. Even if they are receptive, a cluster may lack the ability to absorb people and technology produced by the university. Clusters that are externally, rather than regionally, organized and oriented may even facilitate the diffusion of university-derived benefits outside the region. The university can produce the seeds of new firms and industries, but the region must offer a fertile climate for them to flourish. The key factors related to the industry cluster are its pattern of organization, market trends, and the life cycle stage of the industry or technology.

i

University-based cluster development is a difficult path that requires commitment, time and patience. The success of a university-based cluster initiative requires more than an active, engaged, high quality university. It is also necessary to have appropriate conditions within the regional industry clusters. Within a region, universities are best able to affect the growth of young, emerging clusters, but it takes a broad commitment of significant university resources across a variety of departments aligned with the needs of the cluster.

This report was prepared using a case study strategy. Quantitative data was collected for universities, metropolitan areas and counties in order to develop a sample frame for the selection of the cases. The literature review identified the criteria that for the selection of cases that represented conditions and outcomes. For each case, the project team collected a variety of secondary data, analysis and reports and conducted 15 to 20 interviews with university and regional representatives for each of the eight cases. The methodology is further explained in Appendix 1 – Case Selection Methodology. A summary of each of the cases is provided in the Case Profiles beginning on page 35.

ii

Introduction

This research explored a set of regions and universities (cases) with different initial conditions and ultimately different outcomes (Table 1). Initial conditions related to geographic location, industrial composition, size, and innovation assets (including universities). The primary outcome is employment growth, but we also considered per capita income levels. We selected several different industry clusters from each region in order to have some overlap of clusters across regions.

Selecting only cases with known best practices or with ideal outcomes can identify common factors that might indicate a connection between the practice and the outcome, but it does not distinguish whether these factors and practices are also present in unsuccessful cases. The cross-case approach employed here addresses this problem. With this approach the factors that are unique to the desired outcome or undesired outcome are the most critical. Factors that are common in relation to the desired outcome, but are also associated with the undesired outcome will represent conditions that may be necessary, but not sufficient to impact the desired outcome.

Table 1: Case Taxonomy Economy is growing1 … University is…

Above US average Close to US average Below US average

Engaged Ann Arbor (University of Michigan)

Morgantown (West Virginia University)

New River Valley (Virginia Tech) Allentown (Lehigh University)

Not engaged Las Cruces (New Mexico State) Tallahassee (Florida State)

Waterloo-Cedar Falls (University of Northern Iowa)

Dayton (Wright State University)

These cases are snapshots and do not explore the full historical development of either the regional economies or the universities. Such history and tradition can be important in understanding current conditions. Furthermore, universities and regions are not monolithic; they are comprised of many parts. In reducing these complex entities for analysis, some elements are emphasized at the exclusion of others, and the complexity of the whole is oversimplified. This approach is particularly evident during discussions of

1 The regions were categorized as “Above US Average” if the employment growth from 1990-2000 was more than two percentage points above the US average, and below if it was less than two percentage points below the US average.

Universities and the Development of Industry Clusters Page 2

how engaged or aligned universities are with local economies, especially when some university units or components are significantly involved with local industry and others are not.

The criteria used to select and evaluate cases are regional in scale and thus they blur significant achievements that may be achieved in a specific town, industry or firm. Using the criteria of regional employment growth, rather than cluster growth, establishes a higher standard of performance, but one that is more meaningful. It is a hollow victory to grow one industry cluster if the rest of the economy is in decline. If local successes are not having regional impact, they are not fully recognized in our analysis. The goal is to assess the impact of universities on regional clusters and economic development, not to denigrate local achievements.

After preliminary case analyses were conducted, refined criteria were developed to explore the level of university engagement. We also examined regional economic growth through a variety of measures and perspectives (Appendix 1 – Case Selection Methodology on page 95). This is presented in a simplified taxonomy in Table 1: Case Taxonomy. The taxonomy oversimplifies the complexities in these cases; the distinctions between them are better depicted in Figure 1.

Figure 1: Case Taxonomy University Engagement and Employment Growth

0%

5%

10%

15%

20%

25%

30%

35%

- 1 2 3 4 5 6 7 8 9 1

University Engagement

Empl

oym

ent G

row

th, 1

990-

2000

0

High GrowthLow Engagement

High GrowthHigh Engagement

Low GrowthHigh Engagement

Low GrowthLow Engagement

Las Cruces MSA

Ann ArborMSA

Dayton-Springfield MSA

TallahasseeMSA

Waterloo-Cedar Falls MSA

Allentown-Bethlehem-Easton

New River Valley

Morgantown MSA

Findings

Universities are increasingly recognized as key partners in state and local development efforts, and in some instances are actively engaged in initiatives to promote cluster-based development. Cluster development occupies a shifting middle ground at the intersection of industrial, technology, and regional policy. Industrial policy seeks to


improve the performance of a specific sector of the economy.2 Technology policy promotes the advancement and diffusion of knowledge and innovation, and in its pure form does not target individual firms.3 There is sufficient overlap between industrial and technology policy, so by and large they can be considered one in the same for most purposes. Regional policy aims to develop the economy or improve the socio-economic condition of geographically-targeted places. As it has been embodied in federal and state programs, technology-based economic development contains elements of all three of these policy domains. Cluster policy is a hybrid of these domains, over which no level of government has clear authority or responsibility (much like regional policy), which provides both opportunity and challenge.

Figure 2: Cluster Policy

Industrial Policy

Technology Policy

Regional Policy

Cluster Development

Industrial Policy

Technology Policy

Regional Policy

Cluster Development

Understanding the role of the university requires an appropriate framework. In what has become known as the cluster diamond, Michael Porter mapped four interactive dimensions that impact cluster competitiveness. These are factor conditions, demand conditions, firm strategy and rivalry, and supporting industries. The manner in which firms compete is key in productivity growth and essential to increasing the standard of living. With competition, the focus of firm strategy shifts to innovation rather than cost reduction by lowering labor costs.4 Market and growth oriented firms can push

2 Michael Storper, “Competitiveness policy options: the technology-regions connection,” Growth and Change 26 (Spring 1995): 285-308. 3 Storper 1995; and Michael H. Best, The New Competitive Advantage: The Renewal of American Industry (Oxford and New York: Oxford University Press, 2001). 4 Michael E. Porter, On Competition, Harvard Business Review Book Series (Boston: Harvard Business School Press, 1998).


innovation in a regional cluster, but there must also be sufficient local demand or sensitivity to external demand that provides innovation pull.

Cluster theory also describes how factors external to the firm impact competitiveness and innovation. It is not just the characteristics of firms that create a truly competitive cluster; there are regional factors external to the firm that matter as well. Universities are one such “regional factor” that impacts all of the dimensions of cluster competitiveness. On the one hand, universities are an asset that increases the quality of inputs and producers, by upgrading human capital and disseminating knowledge. Universities also promote economic diversity. In fact, the key role of the university is not so much to grow the economy, as it is to diversify it by generating new opportunities out of the old. The university is the creative side of economic destruction.

The cluster framework is important for examining the role of universities in cluster development and is critical to the design and understanding of this study. Universities are part of the fabric of relationships within a region. Without the context of the cluster, in which the university is one factor, there is the danger of magnifying the role of the university such that one can mistakenly link any observed effect to a university cause. Working within this framework helps to control for this bias and determine which aspects of the university matter most.

University Factors

There are three dimensions in which universities contribute to their local economies. The first is through purchasing and procurement activities. Numerous economic impact studies have demonstrated the significance of this role in terms of the job and income multipliers generated by these functions of the university. However, this does not represent an economic contribution that is unique to universities; the scale of these impacts may differ from other large institutions and employers but they are not substantively different. The second dimension is the traditional function of universities in expanding human capital through education and training. The impact of universities in this regard is equally well documented and generally agreed upon. The only problem is that when universities upgrade human capital they make it more mobile. People with more education are more likely to move longer distances such as to new states or metropolitan areas, and they do it more for work-related reasons.5 Unless the region has a healthy economy and job market, these graduates will leave. If regions want to maximize the human capital benefits provided by universities, then we have to consider the final aspect of how universities contribute to local economies. Related to their role in education and training, universities are creators of knowledge, sources of innovation and generators of economic development. It is this final role in which universities have the greatest potential to affect economic development.

5 Jason Schachter, Why People Move: Exploring the March 2000 Current Population Survey, U.S. Census Bureau, May 2001, 4.


One of the formalized linkages between universities and industry is the technology transfer process, which is the commercialization of technology created by university researchers. Technology transfer can be defined as “the transfer of the results of basic and applied research to design, development, production, and commercialization of new or improved products, services, or processes.”6 Technology transfer became more formalized as a university function in the late 1970s, and is becoming increasingly important at universities across the country, as a source of revenue, a stimulus to the regional economy, and a method of bringing research into practical use. While technology transfer used to consist mainly of patenting, it now includes licensing, research consortia, industrial extension (technical assistance) programs, industrial-liaison or affiliates programs, spin-off enterprises, research parks, start-up firm incubators, consultant services, and venture-capital funds.7 Tech transfer can also include the spread of knowledge through more informal means, such as meetings between academics and industry professionals.

Figure 3: Models for Organizing University Tech Transfer Activity

Source: Gary Matkin, “Spinning off in the U.S.”, OECD Workshop on Research-based Spin-offs, 8 December 1999; available from http://www.oecd.org/dataoecd/48/17/2370995.ppt; Internet; accessed 28 August 2003.

The structure of technology transfer operations varies from university to university. There are some basic characteristics that distinguish different organizational arrangements. The integrated organization is run by university faculty and is part of a university department. An integrated technology transfer office does not have its own 6 Gary Matkin, “Organizing University Economic Development: Lessons from Continuing Education and Tech Transfer,” in The University’s Role in Economic Development: From Research to Outreach. (Jossey-Bass, June 1997). 7 Matkin 1997, 32.

http://www.oecd.org/dataoecd/48/17/2370995.ppt


administrative space, and is sometimes considered to be a university entity. Non-faculty professionals run peripheral offices supervised by a member of the university’s administration. Unlike the integrated office, the peripheral office has its own administrative space and staff. The subsidiary organization is a separate legal entity, usually a non-profit corporation, in which the university holds equity. The interdependent organization is also a separate legal entity, but the university does not hold equity. In the case of WARF, the Wisconsin Alumni Research Fund, the University of Wisconsin provides WARF with intellectual property rights in exchange for research funding. Independent organizations tend to have a contract or informal arrangement with the university, but the university does not have control and usually does not hold equity in the organization.8

Table 2: Tech Transfer in Case Study Universities Organization of Tech Transfer University

License Income9 Industry R&D Startups10

Peripheral Florida State University 34.7% 0.7% 5

Peripheral Lehigh University 0.4% 24.2% 0

Peripheral University of Northern Iowa 0.8% 0.9% 1

Peripheral West Virginia University 0.1% 10.8% 1

Peripheral Wright State University 0.7% 7.7% 0

Peripheral-Integrated Univ. of Michigan 1.4% 6.9% 15

Subsidiary New Mexico State University 0.0% 4.8% 0

Subsidiary Virginia Tech 0.6% 7.4% 9

Note: Based on Matkin’s (1999) framework, classification of universities by the authors data from AUTM.

While these frameworks can be used to organize and classify the technology transfer function at various universities, as well as to differentiate institutional approaches, very little research has been done on which organizational structure is most effective. Examining the case study universities, there is some variation in the technology-transfer activities of the engaged universities. Technology transfer activities at the case study universities were classified using Matkin’s criteria and classifications.11 The case study universities were distributed into one or more of three of Matkin’s prototypes.

Of the case study universities, only Virginia Tech and New Mexico State follow the subsidiary model, while the others are peripheral organizations in Matkin’s (1997) framework. New Mexico State University is similar to Virginia Tech in that it has a

8 Matkin 1997, 33-35. 9 Average license income 1998-2000 as a percent of average R&D expenditures as reported by AUTM. 10 Average industry R&D 1998-2000 as a percent of average R&D expenditures as reported by AUTM. 11 Gary Matkin, “Spinning off in the U.S.”, OECD Workshop on Research-based Spin-offs, 8 December 1999; available from http://www.oecd.org/dataoecd/48/17/2370995.ppt; Internet; accessed 28 August 2003.

http://www.oecd.org/dataoecd/48/17/2370995.ppt


nonprofit technology transfer corporation, but this has not resulted in higher levels of startups, industrial R&D or licensing income. Virginia Tech has performed well in terms of startups and industry R&D. The University of Michigan differs from the peripheral model somewhat as it operates through satellite technology transfer offices for the College of Engineering and the Medical School; it was therefore classified as a peripheral-integrated hybrid. This structure would seem to add a level of confusion for the community as well as faculty if the commercial product is based on interdisciplinary research that requires coordination across one or more of the offices. However, the University of Michigan has generated more startups than the other universities and is also strong in generating licensing income and industry R&D. Florida State’s technology transfer function is organized as a university administrative office; it has been very successful in generating license income, primarily from its best-selling anti-cancer drug, Taxol, but has had less success in generating startups.

Universities have to respond to a variety of priorities not related to economic development. Chief among these is generally the academic mission. In the technology transfer domain, the obligations of the Bayh-Dole Act may encourage a university to license a technology rather than support a start-up. Furthermore, not all R&D is equally effective at generating spin-offs or other benefits that can be captured by a regional economy. Federally-sponsored research is geared toward national goals, whereas some industries, such as the life sciences, require more time and money to yield commercially viable ventures. This does not mean that technology transfer is unimportant, but that it has to be placed within a context of broad university engagement. It is possible that a pattern might emerge based on a study of all universities, their technology transfer structure and outcomes, but the evidence from the case studies is that the structure of the technology transfer office does not determine a university’s performance in generating economic impact.

Research, policy and practice have focused too narrowly in terms of assessing the ability of the university to spur economic development. Much of the literature and policy debates regarding the role of universities in fueling economic development have focused on the formal processes of technology transfer and the role of the university as a generator of knowledge.12 This focus has resulted in a misplaced emphasis and disregard for critical factors in leveraging university assets for growth. As these case studies demonstrate, the organization and operation of technology transfer does not explain differences in generating regional impact.

What makes an innovative university does not automatically make a high-impact university. From another perspective, what is good for the university does not always benefit the region, and vice versa. The challenge is how to achieve a mutual benefit for the university and the region, which requires a better understanding of the role played by universities. Universities are an excellent resource for transforming the economy through

12 There is another domain of the literature that focuses on the economic impact of universities through their purchasing power and through upgrading workforce skills. While these are not trivial effects, they do not fundamentally transform the local economy and are not the focus of our concern.


the creation of new industries, but the ability of these industries to grow the region is related not to the character of the university, but to the character of the region, the state and of the industry itself. The three factors related to the university are:

• Breadth of involvement

• Strong base of R&D

• Regional alignment

Breadth of Involvement

An interesting model of university-based economic development portrays the university, industry, and government as being intertwined in a DNA-like formation. In this model, pioneered by Loydesdorff and Etzkowitz,13 the knowledge sector plays an important role: “three institutional spheres (public, private, and academic) that formerly operated at arms-length in laissez faire societies, are increasingly interwoven with a spiral pattern of linkages emerging at various stages on the innovation and industrial policy-making processes.”14 The authors describe two cycles that universities must go through in order to become more active players in the innovation process. The first includes making research a part of the academic mission, and the second involves taking on a role in regional economic development, both through research and teaching. Once these changes have taken place, the university has a new organizational structure with “mixed disciplinary departments, interdisciplinary centers, new disciplines, self-generation institution, [and] increased social space.”15 While industry and government remain independent, they also change in ways that make it easier for universities to become primary participants in the innovative process.

Universities are becoming increasingly entrepreneurial and engaged with business and industry. At this point, most research universities have created some kind of technology transfer program or industrial-liaison program to interact with the business sector.16 Economic development has become a more common focus in the mission statements of many universities. The Georgia Institute of Technology which is known for its focus on the local economy, states in its Strategic Plan: “Georgia Tech is a leading center for research and technological development that continually seeks opportunities to advance society and the global economic competitiveness of Georgia…”17 Purdue

13 L. Loydesdorff, and H. Etzkowitz, “The Triple Helix as a model for innovation studies,” Science and Public Policy (1998): 25. 14 Henry Etzkowitz, Gebhardt Webster, and B. Terra, “The future of the university and the university of the future: evolution of ivory tower to entrepreneurial paradigm,” Research Policy 29 (2000): 315. 15 Etzkowitz et al., 329. 16 Diane Rahm, University-Industry R&D Collaboration in the United States, the United Kingdom, and Japan (Dordrecht: Kluwer Academic Publishers, 2000): 45-46. 17 Louis Tornatzky, Paul Waugaman, and Denis Gray, Innovation U.: New University Roles in a Knowledge Economy (Raleigh: Southern Growth Policies Board, 2002).


University announces on its website that it is a “powerful resource for the economic development of Indiana.”18

Every university is engaged with its region or local community in multiple ways and assessing that involvement is a difficult task. Furthermore, universities are a collective of entities (schools, departments, centers and institutes) and individuals that are direct agents of interaction with the external world. Assessing the breadth of involvement across a university requires an aggregation of the involvement of all of these agents. What really distinguishes a university that is highly engaged is that the involvement of these units is broad and complementary, rather than compartmentalized. Universities need to address business and legal issues, workforce education, infrastructure, and industry relationships, as well as technology and R&D capacity, in order to yield regional benefits. The most engaged universities demonstrate these kinds of diverse, integrated commitments across administrative and academic units, including the schools of business, engineering, law, medicine, and public policy.

Lehigh University is highly engaged across the administrative, academic and research units. The President of Lehigh University has staked out an active role that is proclaimed on the website and maintains staff that handle federal, state, and community relations.19 Industry and community relations are not exclusive to this office, as the Office of Corporate and Foundation Relations, the Vice Provost for Research, and the Office of Government Relations are also active with regional industry and community groups. A variety of entities throughout the university provide support to local companies. The Manufacturers Resource Center works with manufacturing companies; the Musser Center for Entrepreneurship and the Small Business Development Center, both affiliated with the College of Business and Economics, mentor local entrepreneurs; and the Ben Franklin Technology Center, an independent nonprofit located on the Lehigh campus, supports technology-based economic development. There is also an incubator operated on the Lehigh campus that has achieved national recognition. Lehigh faculty and staff play leadership roles on boards and committees of all four of these organizations, as well as other community economic development organizations. Students are also able to support regional and urban economic development through the Lehigh CORPS.

West Virginia University has expanded its role in local economic development efforts in connection with the introduction of several new federal facilities. Existing programs in energy research have been supplemented by efforts in forensics and biometrics. The university, state and local development organizations are explicitly targeting areas where they can combine their joint resources to leverage federal investment. In part this degree of coordination is possible because WVU does not have to compete with any other research universities in the state. Therefore, the university is both a key contributor and beneficiary from strategies to develop technology, health

18 Tornatzky et al., 80. 19 “Economic Development Engine,” available from http://www3.lehigh.edu/about/economicdevelopment.asp; Internet; accessed 27 August 2003.

http://www3.lehigh.edu/about/economicdevelopment.asp


sciences and entrepreneurship. The university and several federal research centers are the anchors of the emerging I-79 High-Tech Corridor.

In addition to its many research centers developing new technologies, West Virginia has an EDA center, an Entrepreneurship Center at the College of Business and Economics, the Bureau of Business and Economic Research that monitors the local economy and the nationally recognized Regional Research Institute, as well as several other specialized business and economic policy centers. The Center for Entrepreneurial Studies and Development (CESD), affiliated with the College of Engineering provides services to local firms in the areas of training, operations and business development. CESD also serves as the champion or institutional home for several economic development initiatives including a regional entrepreneur’s forum. There are few regional efforts that don’t connect back to the university.

Virginia Tech has both breadth and depth of involvement with the community and local industry. Virginia Tech’s resources include the office of Corporate and Foundation Relations, VT Connect, the Economic Development Assistance Center (EDAC), the Corporate Research Center, and the Office of Outreach and International Affairs that provides a variety of outreach and continuing education programs. Virginia Tech has more than a hundred research centers and institutes, many of which are regionally active. To access the expertise and resources of these centers, businesses and individuals can tap VT Connect or the Virginia Tech Expertise Database.

The university’s outreach activities are both extensive and unique. Continuing education programs at the university reach as many as 8,000 people annually, nearly 5% of the regional population. One of the unique aspects of outreach is Virginia Tech’s Center for Organizational and Technological Advancement (COTA). Since 1994 COTA has connected the university to Virginia organizations and individuals. COTA provides small grants for fellowships that “…focus university resources on specific realworld problems and areas where university expertise can make a distinct contribution.”20

An important element of Virginia Tech’s engagement with business and the community is the EDAC, which is funded by the Economic Development Administration. The EDAC provides technical assistance throughout southwest Virginia’s planning districts. Virginia Tech’s efforts support a variety of industries, including automotives, aeronautics, polymers, and biotechnology. The university has helped transition parts of Virginia from tobacco and textiles to new industries. There are community-oriented efforts as well, such as the Blacksburg Electronic Village, the construction of a modern conference center and the renovation the Hotel Roanoke.

The engineering and business schools at the University of Michigan work extensively with the automotive and information technology industries. Research is distributed throughout the university and nearly every academic unit conducts research. The university recently made a $100 million investment in the Life Sciences Institute to 20 Virginia Tech Outreach and International Affairs, “What is COTA?”, available from http://www.cota.vt.edu/content/what.html; Internet; accessed 3 November 2003.

http://www.cota.vt.edu/content/what.html


bring an interdisciplinary perspective to life sciences research. The university is active in regional economic development planning, particularly with the Michigan Economic Development Corporation and the Washtenaw Development Council. Collaboration between the university and state government resulted in the Life Sciences Corridor and the Michigan Universities Commercialization Initiative. The business school holds symposia for area businesses. The Wolverine Venture Fund can help find funding for startups or make a direct investment. The Institute of Labor and Industrial Relations provides economic and labor market outlook reports. The Center for Local, State, and Urban Policy examines problems facing states, cities and metropolitan regions. The Program for Research on the Information Economy is focused on the economics of information and information systems. The Institute for Social Research, affiliated with the Ford School of Policy, is one of the oldest and largest policy research centers in the nation.

Strong Base of Research

Regional development interests often encourage universities to do more to create new commercial enterprises, but research about the influence of universities on the formation of new companies has been mixed. Bania, Eberts, and Fogarty have found mixed evidence for the role of the university in the creation of new firms. They studied the relationship between university research and development and the birth of new firms (classifying firms by SIC code), and only found a significant relationship in Electrical and Electronic Engineering.21 However, because their data was from 1976-78, it is possible that if they used more recent data they would draw different conclusions. Many researchers have written about the importance of MIT in the creation of new firms around Route 128 in Boston, as well as Stanford’s role in Silicon Valley’s technology start-ups.

The Association of University Technology Managers tracks a number of indicators of technology transfer activities, one of which is startups. The AUTM Survey documented that in 2000, 454 new startups were reported by universities responding to their survey, up from 344 in 1999.22 The average startup rate for all universities from 1998 – 2000 was one per $69 million in R&D expenditures, but there is no significant correlation between R&D spending and the generation of startups, particularly if the outliers are removed (Figure 4). The relationship between startups and research is a skewed distribution. Only four universities in the U.S. were able to spin-off more than ten firms annually, and all of them spent nearly a half billion dollars in annual R&D. Few universities with annual R&D expenditures of less than $100 million generate more than one or two spin-offs annually.

21 Neil Bania, Randall Eberts, and Michael Fogarty, “Universities and the Startup of New Companies: Can we generalize from route 128 and Silicon Valley?” The Review of Economics and Statistics 75 issue 4 (November, 1993). 22 Association of University Technology Managers, “AUTM Licensing Survey: 2001,” edited by Lori Pressman, Chair.


Figure 4: Annual R&D Spending and Start Ups Annual R&D Spending and Start Ups

$-

$100,000,000

$200,000,000

$300,000,000

$400,000,000

$500,000,000

$600,000,000

$700,000,000

$800,000,000

$900,000,000

$1,000,000,000

- 5.0 10.0 15.0 20.0 25.0

Number of Start Ups

R&

D S

pend

ing

Johns Hopkins

University

Massachusetts Institute of Technology

Source: AUTM data 1998, 1999, 2000.

This relationship holds true for the case study universities as well (Figure 5). A large R&D base is necessary but not sufficient to generate economic impact. Without a large R&D base, even highly engaged universities are not able to exert enough impact to make a difference in a small economy. Lehigh University is highly engaged in the regional economy and regional activities, and it is a world leader in several technology areas, but with its small size the positive spillovers from the university can be overpowered by other events. For example, in 2003 Agere Systems, a maker of communications networking products, announced a shift in market focus and manufacturing operations, moving some of the manufacturing out of the region. (See the Lehigh profile on page 57). The decision by Agere was not made on the basis of university R&D or even on regional characteristics, but on industry conditions and competitive pressures. The consolidation of administrative and research personnel in the Allentown headquarters mitigates, but does not replace the loss of manufacturing jobs. Agere is a major anchor, but the region maintains several thousand jobs in this cluster. The uncertainties around Agere may turn out to be only a temporary psychological setback but the situation illustrates the limits of university leverage on a cluster.


Figure 5: Start Ups and R&D Expenditures, 1998-2000 Start Ups and R&D Expenditures, 1998-2000

0

2

4

6

8

10

12

14

16

Florida StateUniversity

LehighUniversity

New MexicoState University

Univ. ofMichigan

University ofNorthern Iowa

Virginia Tech West VirginiaUniversity

Wright StateUniversity

Source: AUTM, 1998 - 2000

Star

tups

-

100,000,000

200,000,000

300,000,000

400,000,000

500,000,000

600,000,000

R&

D E

xpen

ditu

res

Start ups Average R&D Expenditure

Regional Alignment

If a large R&D base is not sufficient for generating regional industry clusters, then what does matter? The alignment of university assets, skills and expertise with regional industry clusters maximizes the regional benefit. Some regions may have a substantial research presence, but companies in the surrounding region are not able to absorb the resulting technology. In these cases innovation is more likely to flow out of the region. It is also a matter of how the geography of university impact aligns with the boundaries of the region and local industry clusters. If the boundaries of the industry cluster overflow the regional boundaries, the impact of the university will be dispersed.

The case study universities and regions were compared on the basis of their cluster concentration and research intensity. Cluster concentration was examined using data from the Harvard Business School’s Cluster Mapping Project. For each region, the location quotients for cluster employment provide an estimate of the concentration of employment in the region relative to the total cluster employment in the U.S. 23 A value greater than one indicates that the region’s share of employment in an industry exceeds the national share. Research intensity was measured for each university based on how the cluster-related academic R&D expenditures per cluster employee in the region varied from the national average. Using these two measures creates four possibilities of alignment between the university and the cluster, each of which suggests a different

23 A location quotient compares the share of cluster employment in a region with the national share. The formula for calculating the values is: Location Quotient = (Cluster Employment in Region / Total Employment in Region) / (Cluster Employment in Nation / Total Employment in Nation).


strategic need (Table 3). University-Cluster examples from the case studies illustrating these alignment options are presented below.

Table 3: University-Cluster Alignment Research Intensity

Low High

High Cluster is dominant Strategy: Focus on building R&D

University & cluster are aligned Strategy: Focus on efficiency and transfer

Cluster Concentration Low Limited foundation

Strategy: Focus on nothing or everything

University is dominant Strategy: Focus on cluster development

The University of Michigan exemplifies alignment of R&D activity and cluster specialization in several of Ann Arbor’s clusters, particularly in information technology (IT) and the life sciences. The university’s research intensity in the information technology cluster is just above the national average, but the university is actively engaged with the cluster in several areas. A group of University of Michigan professors helped to initiate the IT cluster strategy, referred to as the Ann Arbor IT Zone. These professors aimed to emulate the success of Palo Alto, California, by coordinating industry, government, and academia in developing an IT industry.24 They believed Ann Arbor had the resources to begin such a cluster, and formed the IT Zone in response, creating “a partnership of the University of Michigan, businesses and local government to stimulate and grow the IT industry in the area”.25 The IT Zone holds networking sessions for executives, brings together people working in information technology, and works to bring new IT companies to Ann Arbor. Through the Wolverine Venture Fund and the Zell-Lurie Institute, the business school has extensive interaction with IT companies in the region. The fund works with other venture capitalists in the region and also invests its own funds in companies after conducting due-diligence tests. The Zell-Lurie Institute holds symposia twice a year, showcasing the development of new technologies while bringing together entrepreneurs and venture investors.

Ann Arbor’s IT cluster is still relatively small and the region is only slightly more specialized than the nation in IT, but regional development efforts in Ann Arbor have concentrated on growing the IT industry in recent years. The Michigan Economic Development Council, a public-private partnership that serves as the state’s economic development arm, is also an active participant in the IT strategy. It has a variety of supporting initiatives, including industry parks, an angel investor network, workforce training funds, the Michigan Council for IT Executives, and property tax abatements. On a local level, the Washtenaw Development Council is an active participant and incubated the Ann Arbor IT Zone/Business Accelerator. The Accelerator helps IT companies

24 Ann Arbor IT Zone, “Mission and History,” available from http://www.annarboritzone.org/history.asp; Internet; accessed 27 August 2003. 25 Washtenaw Development Council, “Technology Rankings,” available from http://www.wdc-econdev.com/rankings.html; Internet; accessed 27 August 2003.

http://www.annarboritzone.org/history.asphttp://www.wdc-econdev.com/rankings.htmlhttp://www.wdc-econdev.com/rankings.html


relocate to Ann Arbor/Ypsilanti. The Washtenaw Council also tracks the IT industry in the region; in 1995 they conducted a survey of the IT industry in the region and found 200 firms with 3,500 employees; by 2000 these numbers had increased to 1,000 companies with 20,000 employees.26

Figure 6: University of Michigan Alignment

University of Michigan Alignment

Automotive

Information TechnologyMedical Devices

Motor Driven Products

Biopharmaceuticals

(2.00)

-

2.00

4.00

6.00

8.00

10.00

(200.00) - 200.00 400.00 600.00 800.00 1,000.00 1,200.00 1,400.00

R&D Per Employee, Above or Below US

Empl

oym

ent C

once

ntra

tion

vs. U

S

23,265

1,986

1,530821

244

Source: National Science Foundation, R&D Expenditures at Universities and Colleges, by Science and Engineering Field: FY 2001 (Table B-39), and Cluster Mapping Project, Institute for Strategy and Competitiveness, Harvard Business School (cluster data for 2001).

26 Washtenaw Development Council, “Technology Rankings,” available from http://www.wdc-econdev.com/rankings.html; Internet; accessed 27 August 2003.

http://www.wdc-econdev.com/rankings.htmlhttp://www.wdc-econdev.com/rankings.html


Figure 7: Lehigh Alignment Lehigh Alignment

Biopharmaceuticals


Medical Devices Information Technology

Automotive

-

0.50

1.00

1.50

2.00

2.50

3.00

(100.00) (80.00) (60.00) (40.00) (20.00) - 20.00 40.00 60.00 80.00


Empl

oym

ent C

once

ntra

tion

vs. U

S

2,0194,468

2,345

370

306


Lehigh University and the Lehigh Valley region provide a range of examples with clusters in every quadrant. The Lehigh Valley, like much of Pennsylvania has pursued a deliberate strategy of diversifying the economy to avoid over-dependence on any one sector. As a reflection of this approach, there are five emerging clusters targeted by regional development groups with potential university synergy. These clusters tend to pay above-average wages, require technology or knowledge-based skills and have the potential for high-growth. The emerging clusters are chemical and allied products, health-related companies, technology intensive manufacturing/service companies, engineering management and business consulting firms, and financial and insurance services. These clusters are broadly defined, making it difficult to assess the university’s impact, or to compare the clusters across regions.

The analysis here focuses on biopharmaceuticals and medical devices, which are often considered together as part of the life sciences cluster. The life sciences cluster is an emerging cluster in the Lehigh Valley, however, medical devices is a larger and growing presence in the region than biopharmaceuticals. In biopharmaceuticals, the region faces the challenge of growing the cluster and increasing university R&D. In medical devices the cluster is more established in the region. According to the Cluster Mapping Project at the Harvard Business School’s Institute for Strategy and


Competitiveness, between 1990 and 2001 medical devices was the fifth largest job producer among all traded clusters in the region.27 OraSure, the creator of an oral HIV test, and a client of the Ben Franklin incubator, is one of the region’s most successful spin-outs; it is now the 147th largest employer. B. Braun, a manufacturer of disposable surgical and medical supplies, is the 19th largest employer in the Lehigh Valley.28 The challenge for the medical devices cluster in the region and its alignment with Lehigh is to enhance the university’s R&D in that area. This should be helped by the National Science Foundation’s recent award of $1.38 million to Lehigh University to enhance its bioengineering program.

Promoting the growth of the life sciences cluster is the role of the Life Sciences Greenhouse of Central Pennsylvania (LSGPA), whose mission is to enhance and translate life sciences innovation into economic development in central Pennsylvania. The LSGPA has identified three areas of regional significance that have strong synergies with the expertise and resources of Lehigh:

LSGPA Focus Lehigh University Assets Drug design and delivery systems Biopharmaceutical Technology Institute

Biology Program

Biomedical devices Institute for Biomedical Engineering and Mathematical Biology

Materials Research Center

Center for Optical Technologies

Bionanotechnology Bioengineering Program

Center for Advanced Materials and Nanotechnology

Lehigh is one of seven university participants in the LSGPA and brings a number of high quality resources into the initiative. Lehigh’s Center for Optical Technologies is a national leader in optics research, and has attracted commercial activity to the region. Demonstrating the cross-disciplinary nature of Lehigh, much of this optics research has applications in the life sciences field. Lehigh’s Biopharmaceutical Technology Institute focuses on the improvement of processes in biotechnology and in pharmaceuticals. The university participants include:

• Bucknell University

• Dickinson College

• Pennsylvania State University

27 Cluster Mapping Project, Institute for Strategy and Competitiveness, Harvard Business School. Traded clusters sell products across different economic areas. 28 Lehigh Valley Economic Development Corporation, “Estimated Local Employment at Lehigh Valley Employers (9/23/2003)”, available from http://www.lehighvalley.org/assets/pdfs/econ/Employ_largest.pdf; Internet; accessed 13 October 2003.

http://www.lehighvalley.org/assets/pdfs/econ/Employ_largest.pdf


• Juniata College

• Lehigh University

• Messiah College

• Shippensburg University

The LSGPA has a $10 million gap fund where early stage firms can get up to $350,000 for marketing, business development or building a management team. The LSGPA’s Technology Development Fund provides up to $100,000 of convertible debt for one year, with the goal of moving technology toward commercialization. These funds can be used for prototype, proof of concept and commercial feasibility. The LSGPA is encouraging small-business university collaboration, both explicitly in its guidelines, but also in requirements for matching funds and the ability to meet federal standards for life sciences research that often favor the involvement of established firms or universities. The LSGPA strategy and programs are designed to address the critical needs of the life sciences cluster, building both industry resources as well as enhancing research assets.

Figure 8: Wright State Alignment Wright State Alighment

Aerospace Engines

Aerospace Vehicles and Defense

Automotive



Biopharmaceuticals

(1.00)

-

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

(60.00) (50.00) (40.00) (30.00) (20.00) (10.00) - 10.00


Empl

oym

ent C

once

ntra

tion

vs. U

S

1,810

10,105

32,154

2,185

408

261 1,379



Wright State University’s alignment with various segments of the Dayton economy presents an interesting example of cluster-university interaction. The R&D intensity at Wright State is below the national average with several of the largest clusters, due in part to the size and concentration of these clusters in the Dayton region. Dayton has a diversified economy, although many of the clusters support the two most dominant industries – automotive and aerospace. In fact, even these sectors may be converging. In a recent announcement, Delphi Automotive Systems and the Air Force Research Lab at Wright-Patterson Air Force Base (WPAFB) are collaborating to implement Brake-by-Wire technology in passenger vehicles. Much of the R&D to support this sector is conducted not by Wright State or other universities in the region, but by the Air Force Research Lab. Wright-Patterson Air Force Base is the economic anchor in aerospace, technology and manufacturing, conducting several billion dollars of research annually. Wright State has positioned itself to complement the cluster and not duplicate the specialties of other local institutions. The best example of this may be Wright State’s program in aerospace medicine that combines the resources of the medical school in a unique way with the aerospace cluster.

Figure 9: Florida State University Alignment

Florida State University Alignment


Aerospace Vehicles and Defense

(1.50)

(1.00)

(0.50)

-

0.50

1.00

1.50

(200.00) (100.00) - 100.00 200.00 300.00 400.00 500.00 600.00 700.00


Empl

oym

ent C

once

ntra

tion

vs. U

S

177

175

40



Florida State’s research assets engage a broad range of industries. The university has several assets with expertise in materials science that cut across three emerging clusters – aerospace, information technology and medical devices (Figure 9). Florida State University, in partnership with Florida A&M University and Leon County has been working to attract businesses to the region; Innovation Park was created more than 20 years ago as a University-Industry research park. It currently houses the National Science Foundation’s National High Magnetic Field Laboratory and research centers associated with FSU and Florida A&M University. Another research park, spearheaded by Florida State University, is modeled on the University of North Carolina’s Centennial Park.29

Table 4: Sample FSU Cluster Assets Medical

Devices Information Technology

Aerospace

Advanced Mechanics & Materials Laboratory • Center for Biomedical & Toxicological Research and Waste Management •

Florida Advanced Center for Composite Technologies (FAC2T) • • • Information Use Management and Policy Institute •

Materials Research and Technology, Center for (MARTECH) •

Molecular Biophysics, Institute of •

National High Magnetic Field Laboratory • • Pepper Institute on Aging •

Sensory Research Institute •

Florida State is above the US average for research intensity in medical devices, information technology and aerospace although the clusters employee few people in the region. There are numerous research assets for growing the IT cluster in the region, not just at Florida State, but also at Florida A&M University. The Supercomputer Computations Research Institute and the National High Magnetic Field Laboratory are critical assets, but the region’s industry base may be too small to benefit from these assets. There are three IT firms that have made the Inc. 500 list, but none of these firms has more than 200 employees: Mainline Information Systems, an IBM reseller; ATG Technologies, a provider of voice mail and paging services; and Advanced Systems Design, computer consulting services. In order to attract and grow firms there are tax and training incentives of up to $5,000 per new job paying more than 115% of the average private sector wage in the Tallahassee area.

Florida State University exemplifies the case of the high-quality research university that has no local industry to absorb its research. There are only 134

29 Melanie Yeager, “FSU Eyes Proposed Research Park”, The Tallahassee Democrat, 12 June 2002 [journal online] http://www.tallahassee.com/mld/democrat/3449621.htm; Internet; accessed 4 September 2003

http://www.tallahassee.com/mld/democrat/3449621.htm


technology firms in the Tallahassee region and most of those are very small. The Tallahassee economy is also diversified. Of the four clusters that are relatively concentrated in the region – communications equipment, business services, printing and publishing, and furniture, none is significantly more concentrated in Tallahassee than it is nationally.30 Tallahassee is a state capital; therefore much of the metropolitan economy is oriented to government services. With the exception of communications equipment, the clusters in which the region specializes are not clusters that benefit from university research.

Table 5: Technology Firms in Tallahassee Number of Employees Number of Firms

Less than 10 79

11 to 50 40

51 to 100 10

101+ 5

TOTAL 134

Source: Florida State University. High Technology Business List: Tallahassee Region. October 2003.

Cluster Factors

The university dimension is only half of the equation for successful university-industry clusters. The other dimension is the nature and organization of the cluster. Clusters vary based on their life cycle and industry structure as well as their specific pattern of organization in different regions. A cluster may be organized and operate very differently from one region to another. Because cluster organization is critical to the ability of a university to impact cluster dynamics and growth, a review of these dynamics is in order.

There are several explanations for the de-concentration of production that affects cluster growth and decline. These explanations center on either location factors or externalities from the concentration of production (diseconomies of agglomeration). Favoring the de-concentration of production is the need to locate near expanding external markets or to access a dispersed, international labor force. Congestion and high land rents also counteract degree of concentration.31 The difficulty with these explanations is that it is difficult to identify the thresholds when location or agglomeration factors become negative.

The product life cycle offers an alternative explanation for the de-concentration of production that affects regional clusters. Product life cycle theory examines the role of

30 Cluster Mapping Project, Institute for Strategy and Competitiveness, Harvard Business School 31 Paul Krugman, “What's new about the new economic geography?” Oxford Review of Economic Policy V14 2 1998: 7-17.


technological progress from invention to decline in firms or products. Products pass through four phases in the life cycle - invention, growth, maturity, and decline. Each phase is characterized by its own growth trajectory, market structure, factor input, competitive environment, business size, and locational features. The theory emphasizes the growing importance of cost reduction and standardization of products and production processes, and the declining importance of R&D and production flexibility in the mature phase of a product. To remain competitive, firms often shift production from the location of innovation to lower-cost locations (domestic and foreign) to take advantage of lower land, labor, and/or resource costs.32

Audretsch and Feldman extended the product life cycle beyond the question of “who innovates” and “how much innovation takes place” by adding a geographic component describing how regional industry concentrations form as a result of innovation. Controlling for the geographic concentration of production, and using industry R&D, university R&D, and the share of the labor force accounted for by skilled workers as three measures of innovation, they found that production tends to be more “geographically concentrated where new economic knowledge plays an important role,” and is “shaped by the stage of the industry life cycle.” Innovative activity clusters during the innovation and growth stages, and disperses during maturity and decline. 33

Some regions also lose the ability to absorb innovation. As a disruptive new technology is introduced, established clusters in older cities and regions may initially be reluctant to abandon established technology and production methods more profitable than adopting new technology. As new technology becomes more competitive, however, the cities and regions adopting new technologies are able to overtake those that aren’t.34 The steel industry and the competition between integrated production facilities and newer mini-mills have been offered as an example to illustrate this process.

In established regions, cluster institutions can also become victim of lock-in. Research labs, as well as educational and financial institutions, can become captive to existing customers and ways of doing things. Institutions used to serving a dominant industry may be unresponsive to changes in that industry or even ignore industries and opportunities presented by new innovations.35 While regions may find it difficult to unlearn tried-and-true ways of business, competition may play a role in preventing regional lock-in. 36 Certain structures, such as oligopolies, may prevent a region from competing its way out of lock-in. 37

32 Rolf Sternberg, “Regional Growth Theories and High-Tech Regions,” International Journal of Urban and Regional Research 20 (September 1996): 518-38. 33 David B. Audretsch and Maryann P. Feldman, “Innovative Clusters and the Industry Life Cycle,” Review of Industrial Organization 11 no.2 (1996). 34 Elise S. Brezis and Paul Krugman, “Technology and the Life Cycle of Cities,” Journal of Economic Growth 2 no.4 (1997): 369. 35 Peter Maskell, “Towards a knowledge based theory of the geographical cluster,” Industrial and Corporate Change 10 (December 2001): 921-943; Chinitz, 1961. 36 Anders Malmberg and Peter Maskell, “The elusive concept of localization economies: towards a knowledge-based theory of spatial clustering,” Environment and Planning 34 n3 (March 2002): 429-229;


Regions do not have to wait for economic restructuring to alter their economy. They can prepare by seeding new industries to complement or even replace existing ones. Michigan remains heavily reliant on the automotive sector, but it is preparing for a future that may radically alter that industry. Michigan’s NextEnergy initiative is dedicated to developing fuel cell technology that could be incorporated into automobiles, or become a new industry in its own right. Ann Arbor is also working to diversify and develop the regional economy to avoid reliance on one critical industry. Michigan, and Ann Arbor in particular, are focused on life sciences and information technology. Through the year 2000, these efforts have kept the Ann Arbor economy in balance, with strong gains in some sectors offsetting the declines in a few others.

Clusters are also organized and structured differently across regions. How a cluster is structured within a particular region helps to determine the ability of a university to impact the cluster in ways that generate regional benefit. There are four basic cluster development patterns, each of which has different implications for regional growth and development:38

• Networked (Industrial Districts)

• Hub and Spoke

• Satellite (Branch Plant)

• Institutional (State-Anchored)

Examples of each of these four clusters were identified in our case study regions. In some cases a cluster does not fit easily into one category or another, or it may shift over time. For example, the information technology cluster in the Lehigh Valley has the characteristics of both a networked cluster and a hub and spoke cluster. In software, the region is more like a networked cluster with many small firms, but in hardware it is more like a hub and spoke cluster.

David B. Audretsch, “Agglomeration and the Location of Innovative Activity,” Oxford Review of Economic Policy 14 n2 (Summer 1998): 18-29. 37 Walter Adams and Hans Mueller, “The Structure of American Industry” in The Steel Industry (McMillan Publishing Co., 1990): 72-100. 38 Ann Markusen, “Sticky places in slippery space: A typology of industrial districts,” Economic Geography 72 n3 (July 1996): 293-313.


Table 6: Cluster Typology and Cases Networked

(Industrial Districts) Hub & Spoke Satellite

(Branch Plant) Institutional (State-Anchored)

Aerospace Engines & Aerospace Vehicles and Defense

Las Cruces Dayton

Automotive & Motor Driven Products

Ann Arbor Lehigh Valley

New River Valley

Information Technology Ann Arbor Dayton Lehigh Valley

Lehigh Valley

Dayton

Biopharmaceuticals & Medical Devices

Ann Arbor Morgantown Lehigh Valley

Morgantown

Networked Cluster (Industrial Districts)

The prototype is the networked cluster that is composed of networks of small firms in the same or related industries able to rapidly adapt to changing markets and differentiated demand through collaboration and the use of new technologies. Firms in a networked cluster enjoy advantages not available to firms elsewhere, including access to local knowledge and labor markets, low transportation and transaction costs, cultures of flexibility, trust, and cooperation, and available local infrastructure supporting specialized sales, service, and supplier networks. Three frequently cited examples of the networked cluster include Silicon Valley, Boston (Route 128), and northern Italy.39

Figure 10: Networked Clusters

Region

Suppliers

Source: Adapted from Markusen 1996.

39 Markusen, 1996


Michigan is a state with an emerging life sciences cluster, and Ann Arbor is an anchor of that cluster. Michigan has a mixture of small and large firms in life sciences, with most of the larger firms clustered in Western Michigan and the smaller firms in Southeastern Michigan.40 Beyond the region’s impressive research assets, anchored by the University of Michigan, there is a diverse base of firms across several life science sectors. The life sciences cluster in Ann Arbor is not organized around any one entity or even one product or technology area. Rather it is a diverse sector that spans pharmaceuticals, biomedical devices, surgical instruments, medical imaging, diagnostics and therapies. Parke-Davis, maker of Lipitor, was located in the region before Pfizer acquired it and established the Pfizer Global Research and Development facility in the region. Both the University of Michigan and Pfizer/Parke-Davis have been important sources of startup firms for the region’s life science cluster.

Ann Arbor’s efforts to develop the life sciences industry benefit from strong state involvement. The Michigan Life Sciences Corridor is a statewide initiative conceived in 1999 that has broad participation across the state. The University of Michigan is a main participant in the strategy. In addition to providing lab space to companies and training graduates to work in life sciences firms across the state, the University assists with regional infrastructure to support firms. Part of the Corridor strategy involves five labs in Michigan, which provide a variety of services for particular types of life sciences companies. These labs provide small business with access to costly equipment they could not afford on their own. Two of these five labs (Proteomics and Bioinformatics) are at the University of Michigan. The President of the University of Michigan also serves on the Board of the Life sciences Steering Committee at the MEDC.

The university also works with regional economic development organizations in support of life sciences initiatives. Michbio, a trade association devoted to growing the life sciences industry, is located in Ann Arbor. Michbio supports life sciences research and commercialization on a variety of levels, including workforce development initiatives, networking events, and provider services for members. The University of Michigan has been an active partner since the founding of Michbio, providing space for companies on its campus and helping to attract life sciences firms to the region. The Washtenaw Development Council and the University worked together to bring Pfizer’s $250 million R&D plant to Ann Arbor, a major triumph for life sciences in the Ann Arbor region. The university has also invested heavily into the life sciences initiative and set up the Life Sciences Institute on campus with the hope of creating corporate partnerships and promoting commercialization of research. The university has been able to have a large impact both because the cluster is small and innovation- dependent, where the university can exert the most influence.

Dayton’s information technology cluster has aspects of both a networked cluster and an institutional cluster. Wright Patterson Air Force Base is a source of contracts and research producer in information technology, but the IT cluster as a whole is not

40 Catherine D. Freiman, “Ready for the Next Leap Forward: A Competitive Assessment and Strategic Plan to Develop Michigan’s Life Sciences Industry,” Michigan Economic Development Corporation, 2003.


dependent on contract work from the base or its laboratory. Major firms, such as LexisNexis, Reynolds & Reynolds and the Uniform Code Council, balance the influence of WPAFB in the cluster. Most IT firms in the Dayton area have five or fewer employees and are not dependent on WPAFB or the major firms.

In fact, one regional program, the I-Zone, is focused on creating a supportive climate for innovation in Dayton. The I-Zone’s Pillar program is a unique effort designed to help small technology firms sell goods and services to large regional firms. This is an explicit effort to build more linkages between firms within the cluster that should strengthen the network of IT firms. The Greater Dayton IT Alliance, another networking resource, is a 200 member trade association that provides a variety of member benefits and programs and supports the Dayton-Metro Internet Exchange (D-MIX), a local peering point to increase the IT connectivity in the community.

Relative to the size of the cluster, Wright State has intensive research assets supporting information technology. Wright State University, the Dayton Development Coalition and several other groups are promoting the development of information technology, but employment in this cluster remains small. Wright State was the main academic institution sponsoring the Dayton proposal for the Wright Center of Innovation for Advanced Data Management and Analysis, a collaboration with numerous large firms around the state. If funded, this Center would boost the R&D intensity for this cluster.

Hub and Spoke

The hub-and-spoke cluster is characterized by large anchor firms whose suppliers and service providers often concentrate around them like spokes on a wheel. Smaller firms in the region can be closely linked to the dominant firm through supply chains, or may simply be located nearby to take advantage of the benefits of agglomeration. Unlike networked clusters, large firms dominate the inter-firm relationships. These interactions are based more on supply linkages, and less on collaborative innovation sharing. Financial and business services are tailored to the needs of dominant firms, and labor markets are less flexible than in the networked cluster.41 The fortunes of the region are thus tied directly to the dominant firm or industry, which can inhibit a region’s ability to embrace new opportunities.42

41 Markusen, 1996 42 Benjamin Chinitz, “Contrasts in agglomeration: New York and Pittsburgh,” in Papers and Proceedings of the Seventy-third annual meeting of the American Economic Association, The American Economic Review 51 no.2 (May 1961): 279-289.


Figure 11: Hub and Spoke Clusters

Region


Satellite (Branch Plant)

There is a flipside to the Hub and Spoke cluster. The region that is home to the headquarters may be the hub, but other regions are home only to the branch plants. The clusters that are spokes to an external hub may be called satellite clusters. Minimal trade or cooperation takes place within the region, as most linkages are to external supply chains and other facilities of the parent corporation. Labor markets are usually internal to the firm, and characterized by high degrees of labor migration to and from the district in the higher levels of the market.43

Figure 12: Satellite Clusters

Region


Documents

Universities and the Development of Industry ClustersExecutive Summary Regional efforts to develop industry clusters increasingly include universities as central assets. Unfortunately,